The goal of this project is to develop a novel differential phase contrast cone beam computed tomography (DPC-CBCT) technique into a clinically useful diagnostic breast-imaging tool. Potentially, this novel imaging technique will provide true 3D images of breast anatomy in vivo with spatial resolution comparable to that of pathologic images while maintaining patient dose to that of diagnostic mammography. Absorption-based x-ray imaging techniques have shown their limitations in breast imaging because of low absorption contrast difference among breast tissues. One solution for this issue is the emerging phase- contrast technique that is expected to significantly enhance tissue contrast with high spatial and contrast resolution at comparable or even lower dose level than current x-ray imaging methods. Various phase-contrast techniques have been developed in the past decade, among which a recently reported grating-based method, called differential phase-contrast (DPC) technique, will be used for this project. The major advantage of DPC is that a hospital-grade x-ray tube can be used, while for other phase-contrast techniques synchrotron radiation or a micro-focus x-ray tube is required. These are not practical for widespread clinical applications because of either limited availability or extremely low output x-ray power. This project's goal is to incorporate a differential phase contrast (DPC)-based cone beam computed tomography (DPC-CBCT) imaging system into current cone-beam breast CT (CBBCT) scanner as a hybrid system, to perform volume-of-interest (VOI) imaging of any suspicious mass found by the CBBCT scanner for breast cancer diagnosis. It is anticipated that a spatial resolution of up to 25 lp/mm and a tissue contrast of over a thousand times larger than that of absorption-based images can be achieved in the tomographic reconstruction slices with the proposed DPC-CBCT, and thus it should be capable of revealing pathology-level details for more accurate characterization and diagnosis of breast cancers. When combined with the current CBBCT system, it is expected to reduce the breast biopsy rate and accurately detect breast cancer margins. The specific aims of this project include: 1) Perform computer simulation to optimize the parameters for the design and construction of the table-top DPC system, 2) Develop phase retrieval algorithms and reconstruction algorithms for VOI DPC-CBCT imaging, 3) Design and construct a table-top DPC-CBCT system, 4) Evaluate the table-top DPC-CBCT to use table-top experimental data as design inputs for the hybrid system, 5) Design and construct a hybrid breast CT (HBCT) system combining a current CBBCT scanner with a DPC- CBCT system, 6) Characterize the hybrid system with phantoms and specimens, 7)Perform small animal studies in vivo to validate HBCT, and 8) Perform a patient pilot study on the HBCT system.